Right now there's a problem, while implementing Hume on mobile and handling network drops, that if the web socket breaks client side and the user tries to reconnect to wss://api.hume.ai/v0/evi/chat with a resumed_chat_group_id to resume their chat, the web socket is closed with an error "Cannot resume chat group x because it already contains an active chat.".

Takes like a minute for server to time out the active chat.

There should be a way to connect back without having to wait for the server to time out the broken connection.

I assume this was implemented to guard against multiple clients using the resumed_chat_group_id, but then maybe there could be another ID, like an uuid generated in memory just for that single device, so the server knows it's safe to override the active chat, but keep dropping every other device, if it sees someone connecting with that unique uuid.